
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from os import walk, getcwd

In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The text.latex.preview rcparam was deprecated in Matplotlib 3.3 and will be removed two minor releases later.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The mathtext.fallback_to_cm rcparam was deprecated in Matplotlib 3.3 and will be removed two minor releases later.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: Support for setting the 'mathtext.fallback_to_cm' rcParam is deprecated since 3.3 and will be removed two minor releases later; use 'mathtext.fallback : 'cm' instead.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The validate_bool_maybe_none function was deprecated in Matplotlib 3.3 and will be removed two minor releases later.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The savefig.jpeg_quality rcparam was deprecated in Matplotlib 3.3 and will be removed two minor releases later.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The keymap.all_axes rcparam was deprecated in Matplotlib 3.3 and will be removed two minor releases later.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The animation.avconv_path rcparam was deprecated in Matplotlib 3.3 and will be removed two minor releases later.
In C:\ProgramData\Anaconda3\lib\site-packages\matplotlib\mpl-data\stylelib\_classic_test.mplstyle: 
The animation.avconv_args rcparam was deprecated in Matplotlib 3.3 and will be removed two minor releases later.


f = []
for (dirpath, dirnames, filenames) in walk(getcwd()):
    f.extend(filenames)
    break


import pyreadstat


def read_spss(module, f = f):
    files = [i for i in f if module in i]
    if len(files) == 1:
        return pyreadstat.read_sav(files[0], apply_value_formats=True, 
                                formats_as_category=True, formats_as_ordered_category=False)
    else: raise Exception('More than one file!')


ac, ac_meta = read_spss('ac')
#mh, mh_meta = read_spss('mh')
ph, ph_meta = read_spss('ph')
#dn, dn_meta = read_spss('dn')
ep, ep_meta = read_spss('ep')
#co, co_meta = read_spss('co')
gv_imputations, gv_imputations_meta = read_spss('gv_imputations')


def get_desc_df(df, meta):
    desc = pd.DataFrame({
        "var":meta.column_names,
        "varlab":meta.column_labels,
        "values":list(meta.readstat_variable_types.values()),
        "uniques":[len(set(df[i])) for i in df],
        'nas': list(df.isna().sum()),

    })
    return desc


ac_desc = get_desc_df(ac, ac_meta)
gov_dsec = get_desc_df(gv_imputations, gv_imputations_meta)


cols_to_use = [i for i in ac if (('ac036_' not in i) and ('ac037_' not in i) and ('ac038_' not in i) and ('ac040' not in i)
                                and ('hhid7' not in i) and ('mergeidp7' not in i) and ('coupleid7' not in i) and ('language' not in i))]


ac = ac[cols_to_use]


ac[ac['country'] == 'Romania'].isna().sum()

mergeid        0
country        0
ac012_         5
ac014_         5
ac015_         5
ac016_         5
ac017_         5
ac018_         5
ac019_         5
ac020_         5
ac021_      2114
ac022_         5
ac023_         5
ac024_         5
ac025_         5
ac035d1       43
ac035d4       43
ac035d5       43
ac035d7       43
ac035d8       43
ac035d9       43
ac035d10      43
ac035dno      43
ac701_         5
ac702_         5
ac703_         5
ac704_         5
ac705_         5
ac706_         5
ac707_         5
ac708_         5
ac709_         5
ac710_         5
ac711_         5
ac740_         5
dtype: int64


ac['ac021_'].cat.add_categories(['Interpolated'], inplace = True)


ac = ac.apply(
    lambda x: x.apply(
        lambda y: np.nan if (y == "Don't know") or (y == "Refusal") else y
                        
    )
)


ac.loc[ac['country'] == 'Romania', ['ac021_']] = 'Interpolated'


ac[ac['country'] == 'Romania'].isna().sum()

mergeid      0
country      0
ac012_      14
ac014_       7
ac015_       8
ac016_       8
ac017_       7
ac018_       8
ac019_       7
ac020_       7
ac021_       0
ac022_       7
ac023_       7
ac024_       7
ac025_       8
ac035d1     44
ac035d4     44
ac035d5     44
ac035d7     44
ac035d8     44
ac035d9     44
ac035d10    44
ac035dno    44
ac701_       7
ac702_       7
ac703_       7
ac704_       7
ac705_       7
ac706_       7
ac707_       7
ac708_       8
ac709_       7
ac710_       8
ac711_       7
ac740_       5
dtype: int64


ac = ac.dropna().reset_index(drop=True)


for i in ac:
  
    ac[i] = ac[i].apply(lambda x: 1 if x == 'Not selected' else x)\
    .apply(lambda x: 5 if x == 'Selected' else x)


    ac[i] = ac[i].apply(lambda x: 1 if x == 'Never' else x)\
    .apply(lambda x: 2 if x == 'Rarely' else x)\
    .apply(lambda x: 3 if x == 'Sometimes' else x)\
    .apply(lambda x: 4 if x == 'Often' else x)


    ac[i] = ac[i].apply(lambda x: 1 if x == 'Disagree strongly' else x)\
    .apply(lambda x: 2 if x == 'Disagree a little' else x)\
    .apply(lambda x: 3 if x == 'Neither agree nor disagree' else x)\
    .apply(lambda x: 4 if x == 'Agree a little' else x)\
    .apply(lambda x: 5 if x == 'Agree strongly' else x)


gv_aim = gv_imputations[['mergeid', 'age', 'gender', 'single', 'mstat', 'yedu', 'cjs', 'thinc2']].dropna()


gv_aim.shape

(386010, 8)


gv_aim = gv_aim.groupby('mergeid').first()


gv_aim = gv_aim[gv_aim['age'] >= 50]


df = ac.merge(gv_aim, on='mergeid')


df_wage_life_satisfaction = df.groupby('country').agg({"thinc2":"median", "ac012_":"mean"})
df_wage_life_satisfaction.columns = ['Median Wage', 'Life Satisfaction']


df_wage_life_satisfaction


import plotly.express as px


fig = px.choropleth(df_wage_life_satisfaction.reset_index(), locations="country", 
                    locationmode='country names', color="Median Wage", 
                    hover_name="country", scope ='europe', color_continuous_scale="blues", 
                    title='Median Wage')
fig.update(layout_coloraxis_showscale=True)
fig.show()


fig = px.choropleth(df_wage_life_satisfaction.reset_index(), locations="country", 
                    locationmode='country names', color="Life Satisfaction", 
                    hover_name="country", scope ='europe',color_continuous_scale="greens", 
                    title='Life Satisfaction')
fig.update(layout_coloraxis_showscale=True)
fig.show()


es = df.groupby('cjs').agg({"ac012_":"mean", "ac016_":"mean", "ac017_":"mean"})
es.columns = ['LifeSat', 'LeftOut', 'DoWhatWant']


es.dropna().plot.barh()

<AxesSubplot:ylabel='cjs'>


df = df.apply(
    lambda x: x.apply(
        lambda y: np.nan if (y == "Interpolated") else y
                        
    )
)


df = df.dropna()


corr_df = df[['ac014_', 'ac015_', 'ac016_', 'ac017_', 'ac018_',
              'ac019_', 'ac020_', 'ac021_', 'ac022_', 'ac023_',
              'ac024_', 'ac025_']]


corr = corr_df.corr()
corr.style.background_gradient(cmap='coolwarm')


corr_series = corr_df.corr().unstack().sort_values().drop_duplicates()


corr_series[corr_series > 0.3]

ac017_  ac021_    0.321422
ac022_  ac023_    0.331900
ac017_  ac024_    0.348815
ac025_  ac017_    0.350667
ac022_  ac020_    0.361880
ac023_  ac017_    0.366042
ac024_  ac022_    0.383088
ac023_  ac020_    0.384247
ac014_  ac016_    0.386468
ac025_  ac022_    0.402166
ac024_  ac020_    0.413541
ac025_  ac020_    0.437354
ac023_  ac021_    0.444719
ac022_  ac021_    0.450083
ac015_  ac014_    0.452976
ac024_  ac021_    0.474275
ac025_  ac021_    0.500598
ac016_  ac015_    0.503004
ac021_  ac020_    0.553536
ac025_  ac023_    0.557246
ac023_  ac024_    0.562482
ac025_  ac024_    0.643846
ac014_  ac014_    1.000000
dtype: float64


corr_series[corr_series < -0.3]

ac014_  ac023_   -0.416880
ac025_  ac014_   -0.364573
ac016_  ac025_   -0.354189
ac015_  ac023_   -0.351900
ac024_  ac016_   -0.345645
ac014_  ac024_   -0.344952
ac023_  ac016_   -0.341738
ac015_  ac025_   -0.329618
ac021_  ac016_   -0.322571
dtype: float64


%load_ext rpy2.ipython

C:\Users\kiril\AppData\Roaming\Python\Python38\site-packages\rpy2\robjects\packages.py:366: UserWarning:

The symbol 'quartz' is not in this R namespace/package.


# %%R -i corr_df
# install.packages("dendextend")
# library(dendextend)
# aac = cor(corr_df)
# scree(aac)


import pandas as pd
from sklearn.datasets import load_iris
from factor_analyzer import FactorAnalyzer
import matplotlib.pyplot as plt


fa = FactorAnalyzer()
fa.fit(corr_df)
eigen_values, vectors = fa.get_eigenvalues()


plt.plot(vectors+1,'-.')
plt.title('Scree Plot')
plt.xlabel('component number')
plt.ylabel('Eigen values')

Text(0, 0.5, 'Eigen values')


eigen_values

array([4.47569495, 1.38995794, 0.99747547, 0.79102704, 0.75501688,
       0.6843992 , 0.63857954, 0.57629819, 0.47843893, 0.43596334,
       0.42505564, 0.35209289])


fa.set_params(n_factors=2)
fa.fit(corr_df)
loadings = fa.loadings_


factors = fa.transform(corr_df)


factors_df = pd.DataFrame(factors)


corr_df['Factor_1'] = factors_df[0]
corr_df['Factor_2'] = factors_df[1]

<ipython-input-45-501b418dab6d>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-45-501b418dab6d>:2: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


corr = corr_df.corr()
corr.style.background_gradient(cmap='coolwarm')


corr[corr['Factor_1'] > 0]['Factor_1'].sort_values(ascending=False)

Factor_1    1.000000
ac025_      0.761609
ac024_      0.740141
ac021_      0.716329
ac023_      0.677795
ac020_      0.621347
ac022_      0.547563
ac017_      0.473680
Name: Factor_1, dtype: float64


corr[corr['Factor_2'] > 0]['Factor_2'].sort_values(ascending=False)

Factor_2    1.000000
ac015_      0.746140
ac016_      0.712579
ac014_      0.640906
ac019_      0.447540
ac018_      0.316507
Name: Factor_2, dtype: float64


factors_df['country'] = df['country']


factors_df.columns = ['MR1', 'MR2', 'country']


factors_df.groupby('country').mean()


fig = px.choropleth(factors_df.groupby('country').mean().reset_index(), locations="country", 
                    locationmode='country names', color="MR1", 
                    hover_name="country", scope ='europe', color_continuous_scale="blues", 
                    title='MR1mean')
fig.update(layout_coloraxis_showscale=True)
fig.show()


fig = px.choropleth(factors_df.groupby('country').median().reset_index(), locations="country", 
                    locationmode='country names', color="MR1", 
                    hover_name="country", scope ='europe', color_continuous_scale="blues", 
                    title='MR1median')
fig.update(layout_coloraxis_showscale=True)
fig.show()


fig = px.choropleth(factors_df.groupby('country').mean().reset_index(), locations="country", 
                    locationmode='country names', color="MR2", 
                    hover_name="country", scope ='europe', color_continuous_scale="blues", 
                    title='MR2mean')
fig.update(layout_coloraxis_showscale=True)
fig.show()


fig = px.choropleth(factors_df.groupby('country').median().reset_index(), locations="country", 
                    locationmode='country names', color="MR2", 
                    hover_name="country", scope ='europe', color_continuous_scale="blues", 
                    title='MR2median')
fig.update(layout_coloraxis_showscale=True)
fig.show()


wq, wq_meta = read_spss('wq')


wq_desc = get_desc_df(wq, wq_meta)


wq_desc[wq_desc['varlab'].str.contains('press')]


wq[[i for i in wq if 'wq018' in i]].info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 77202 entries, 0 to 77201
Data columns (total 17 columns):
 #   Column    Non-Null Count  Dtype   
---  ------    --------------  -----   
 0   wq018_1   4647 non-null   category
 1   wq018_2   4447 non-null   category
 2   wq018_3   3138 non-null   category
 3   wq018_4   2082 non-null   category
 4   wq018_5   1216 non-null   category
 5   wq018_6   760 non-null    category
 6   wq018_7   430 non-null    category
 7   wq018_8   263 non-null    category
 8   wq018_9   132 non-null    category
 9   wq018_10  73 non-null     category
 10  wq018_11  34 non-null     category
 11  wq018_12  23 non-null     category
 12  wq018_13  12 non-null     category
 13  wq018_14  7 non-null      category
 14  wq018_15  5 non-null      category
 15  wq018_16  1 non-null      category
 16  wq018_18  1 non-null      category
dtypes: category(17)
memory usage: 1.3 MB


wq['wq004_'].value_counts()

Agree                12994
Disagree             12898
Strongly agree        6443
Strongly disagree     5152
Don't know             929
Refusal                 21
Name: wq004_, dtype: int64


wq = wq.apply(
    lambda x: x.apply(
        lambda y: np.nan if (y == "Don't know") or (y == "Refusal") else y
                        
    )
)
wq['wq004_'].value_counts()

Agree                12994
Disagree             12898
Strongly agree        6443
Strongly disagree     5152
Name: wq004_, dtype: int64


poor_jq_prep = wq[['mergeid', 'wq004_']]


poor_jq_prep.columns = ['mergeid', 'pressure']


wq_desc[wq_desc['varlab'].str.contains('emo')]


poor_jq_prep['emo_demanding'] = wq['wq005_']

<ipython-input-65-1dfa862f0fc4>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


wq_desc[wq_desc['varlab'].str.contains('confl')]


poor_jq_prep['conflicts'] = wq['wq006_']

<ipython-input-67-1aba32a59121>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


wq_desc[wq_desc['var'].str.contains('wq00')]


poor_jq_prep['phys_demanding'] = wq['wq002_']
poor_jq_prep['uncomfortable'] = wq['wq003_']

<ipython-input-69-dea2890a9b43>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-69-dea2890a9b43>:2: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


poor_jq_prep['security'] = wq['wq014_']
poor_jq_prep['negative_freedom'] = wq['wq007_']
poor_jq_prep['skills'] = wq['wq008_']
poor_jq_prep['support'] = wq['wq011_']
poor_jq_prep['recognition'] = wq['wq009_']

<ipython-input-70-aabcd66dbf26>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-70-aabcd66dbf26>:2: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-70-aabcd66dbf26>:3: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-70-aabcd66dbf26>:4: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

<ipython-input-70-aabcd66dbf26>:5: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


poor_jq_prep = poor_jq_prep.apply(
    lambda x: x.apply(
        lambda y: np.nan if (y == "Don't know") or (y == "Refusal") else y
                        
    )
)


poor_jq_prep.dropna(inplace = True)


len(set(poor_jq_prep['mergeid'])) == len(poor_jq_prep['mergeid'])

True


for i in poor_jq_prep:
    if i == 'mergeid':
        continue
    print(i)
    print(set(poor_jq_prep[i]))

pressure
{'Strongly agree', 'Agree', 'Strongly disagree', 'Disagree'}
emo_demanding
{'Strongly agree', 'Agree', 'Strongly disagree', 'Disagree'}
conflicts
{'Strongly agree', 'Agree', 'Strongly disagree', 'Disagree'}
phys_demanding
{'Agree', 'Strongly agree', 'Strongly disagree', 'Disagree'}
uncomfortable
{'Strongly agree', 'Agree', 'Strongly disagree', 'Disagree'}
security
{'Agree', 'Strongly agree', 'Strongly disagree', 'Disagree'}
negative_freedom
{'Agree', 'Strongly agree', 'Strongly disagree', 'Disagree'}
skills
{'Strongly agree', 'Agree', 'Strongly disagree', 'Disagree'}
support
{'Strongly agree', 'Agree', 'Strongly disagree', 'Disagree'}
recognition
{'Agree', 'Strongly agree', 'Strongly disagree', 'Disagree'}


for i in poor_jq_prep:
    poor_jq_prep[i] = poor_jq_prep[i].apply(lambda x: 1 if x == 'Strongly disagree' else x)\
    .apply(lambda x: 2 if x == 'Disagree' else x)\
    .apply(lambda x: 3 if x == 'Agree' else x)\
    .apply(lambda x: 4 if x == 'Strongly agree' else x)


for i in poor_jq_prep:
    if i == 'mergeid':
        continue
    print(i)
    print(poor_jq_prep[i].value_counts())

pressure
3    12522
2    12324
4     6233
1     4855
Name: pressure, dtype: int64
emo_demanding
3    12241
2    11528
4     6385
1     5780
Name: emo_demanding, dtype: int64
conflicts
2    16066
1    10059
3     7413
4     2396
Name: conflicts, dtype: int64
phys_demanding
3    11466
4    10012
2     9141
1     5315
Name: phys_demanding, dtype: int64
uncomfortable
2    12154
3     9621
1     8658
4     5501
Name: uncomfortable, dtype: int64
security
3    17811
2     7306
4     6905
1     3912
Name: security, dtype: int64
negative_freedom
2    12817
3    10455
1     7879
4     4783
Name: negative_freedom, dtype: int64
skills
3    16062
4     8980
2     7744
1     3148
Name: skills, dtype: int64
support
3    20102
2     6892
4     6773
1     2167
Name: support, dtype: int64
recognition
3    17925
4     9119
2     6606
1     2284
Name: recognition, dtype: int64


def poor_jq_gen(pressure, emo_demanding, conflicts, phys_demanding, unconfortable,
               freedom, skills, support, recognition, security):
    effort = sum([pressure, emo_demanding, conflicts, phys_demanding, unconfortable])/5
    reward = sum([freedom, skills, support, recognition, security])/5
    return effort/reward

def bad_stress_gen(pressure, emo_demanding, conflicts,
                  freedom, skills, support, recognition, security):
    stress = sum([pressure, emo_demanding, conflicts])/3
    reward = sum([freedom, skills, support, recognition, security])/5
    return stress/reward


poor_jq_prep['freedom'] = poor_jq_prep['negative_freedom'].apply(lambda x: 5-x)


poor_jq_variable = poor_jq_prep.apply(lambda row: poor_jq_gen(row['pressure'], row['emo_demanding'], row['conflicts'], row['phys_demanding'], row['uncomfortable'],
                                           row['freedom'],row['skills'],row['support'],row['recognition'],row['security'])
                   ,axis = 1)

bad_stress_variable = poor_jq_prep.apply(lambda row: bad_stress_gen(row['pressure'], row['emo_demanding'], row['conflicts'],
                                           row['freedom'],row['skills'],row['support'],row['recognition'],row['security'])
                   ,axis = 1)


poor_jq_prep['poor_jq'] = poor_jq_variable
poor_jq_prep['bad_stress'] = bad_stress_variable


wq_desc[wq_desc['varlab'].str.contains('salary')]


wq024 = wq[['mergeid', 'wq010_']]


wq024['wq010_'] = wq024['wq010_'].apply(lambda x: 1 if x == 'Strongly disagree' else x)\
    .apply(lambda x: 2 if x == 'Disagree' else x)\
    .apply(lambda x: 3 if x == 'Agree' else x)\
    .apply(lambda x: 4 if x == 'Strongly agree' else x)

<ipython-input-83-b691e2a40a59>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


wq024.dropna()


df_case3 = wq024.merge(poor_jq_prep, on='mergeid')


df_case3 = df_case3[['mergeid', 'wq010_', 'poor_jq', 'bad_stress']]


df_from_activities_needed = df[['mergeid', 'ac035d1', 'ac035d4', 'ac035d7', 'ac035d8', 'ac035d9', 'ac035d10', 'ac035dno']]


df_case3 = df_case3.merge(df_from_activities_needed, on='mergeid')


ep = ep[ep['mergeid'].isin(df_case3['mergeid'])]


ep_desc = get_desc_df(ep, ep_meta)


ep_desc.to_csv('employment_desc.csv')


ep_desc[ep_desc['varlab'].str.contains('industry')]


ep.shape

(32949, 436)


re, re_meta = pyreadstat.read_sav('sharew7_rel8-0-0_re.sav', apply_value_formats=True, 
                                formats_as_category=True, formats_as_ordered_category=False)


re_desc = get_desc_df(re, re_meta)


re = re[re['mergeid'].isin(df_case3['mergeid'])]


re_desc[re_desc['varlab'].str.contains('industry')]


industry = re['re014_1']


re_desc[re_desc['varlab'].str.contains('empl')]


type_of_employment = re['re015_1']


re_desc[re_desc['varlab'].str.contains('full')]


full_part_time_job = re['re016_1']


employement = pd.DataFrame({
    "mergeid": re['mergeid'],
    "type_of_emp":type_of_employment,
    "full_part":full_part_time_job,
    "industry":industry
})


employement = employement.apply(
    lambda x: x.apply(
        lambda y: np.nan if (y == "Don't know") or (y == "Refusal") else y
                        
    )
)


employement = employement.join(pd.get_dummies(employement[['type_of_emp', 'full_part', 'industry']]))
employement = employement[[i for i in employement.columns if i not in ['type_of_emp', 'full_part', 'industry']]]
employement


df_case3 = df_case3.merge(employement, on='mergeid')


socio_df = gv_imputations[['mergeid', 'age', 'gender', 'yedu']]


socio_df = socio_df.join(pd.get_dummies(socio_df['gender']))


socio_df['age'].plot.hist()
plt.title('Age')

Text(0.5, 1.0, 'Age')


# create a list of our conditions
conditions = [
    ((socio_df['age'] >= 50.0) & (socio_df['age'] <= 54.0)),
    ((socio_df['age'] >= 55.0) & (socio_df['age'] <= 59.0)),
    ((socio_df['age'] >= 60.0) & (socio_df['age'] <= 64.0)),
    ]

# create a list of the values we want to assign for each condition
values = ['age_50_54', 'age_55_59', 'age_60_64']

# create a new column and use np.select
socio_df['age_cat'] = np.select(conditions, values)


socio_df = socio_df[socio_df['age_cat'] != '0']


gv_aim


socio_df = socio_df.merge(gv_aim.reset_index()[['mergeid','thinc2']], on='mergeid')


socio_df['thinc2'].plot.hist()
plt.title('Income')

Text(0.5, 1.0, 'Income')


socio_df[['thinc2']].describe(percentiles=[0.33, 0.66])


# create a list of our conditions
conditions = [
    ((socio_df['thinc2'] >= 0) & (socio_df['thinc2'] <= 12000)),
    ((socio_df['thinc2'] >= 12000) & (socio_df['thinc2'] <= 27600)),
    ((socio_df['thinc2'] >= 27600)),
    ]

# create a list of the values we want to assign for each condition
values = ['income_low', 'income_med', 'income_high']

# create a new column and use np.select
socio_df['Income_cat'] = np.select(conditions, values)


socio_df['yedu'] = socio_df['yedu'].apply(lambda x: np.nan if isinstance(x, str) else x)


socio_df['yedu'].describe(percentiles=[0.33, 0.66])

count    142087.000000
mean         11.997062
std           3.793074
min           1.000000
33%          11.000000
50%          12.000000
66%          13.000000
max          35.000000
Name: yedu, dtype: float64


# create a list of our conditions
conditions = [
    ((socio_df['yedu'] >= 0) & (socio_df['yedu'] <= 11)),
    ((socio_df['yedu'] >= 12) & (socio_df['yedu'] <= 14)),
    ((socio_df['yedu'] > 14)),
    ]

# create a list of the values we want to assign for each condition
values = ['edu_low', 'edu_med', 'edu_high']

# create a new column and use np.select
socio_df['edu_cat'] = np.select(conditions, values)


socio_df['edu_cat'] = socio_df['edu_cat'].apply(lambda x: np.nan if x == '0' else x)


socio_df['edu_cat'].value_counts()

edu_low     59403
edu_med     50036
edu_high    32648
Name: edu_cat, dtype: int64


socio_df['Income_cat'].value_counts()

income_low     50113
income_high    47670
income_med     45048
Name: Income_cat, dtype: int64


socio_df = socio_df.join(pd.get_dummies(socio_df[['edu_cat','Income_cat', 'age_cat']]))


socio_df = socio_df[['mergeid', 
          'Female', 'Male',
          'edu_cat_edu_high', 'edu_cat_edu_low', 'edu_cat_edu_med', 
          'Income_cat_income_high', 'Income_cat_income_low', 'Income_cat_income_med',
         'age_cat_age_50_54', 'age_cat_age_55_59', 'age_cat_age_60_64']]


df_case3 = df_case3.merge(socio_df, on='mergeid')


hh, hh_meta = read_spss('gv_health')


hh = hh.apply(
    lambda x: x.apply(
        lambda y: np.nan if (y == "Don't know") or (y == "Refusal") else y
                        
    )
)


health = hh[['mergeid','gali', 'sphus']]


hh_desc = get_desc_df(hh, hh_meta)


hh_desc[hh_desc['varlab'].str.contains('chronic')]


hh['chronic2w7'].value_counts()

2+ chronic diseases     39451
Less than 2 diseases    37215
Name: chronic2w7, dtype: int64


health['chronic2'] = np.where(hh['chronic2w7']== '2+ chronic diseases', 1, 0)

<ipython-input-132-4c042e694e72>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


health['gali'] = np.where(hh['gali']== 'Limited', 1, 0)

<ipython-input-133-48a2dc68b2fe>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


health['sphus'] = health['sphus'].apply(lambda x: 1 if x == 'Poor' else x)\
    .apply(lambda x: 2 if x == 'Fair' else x)\
    .apply(lambda x: 3 if x == 'Good' else x)\
    .apply(lambda x: 4 if x == 'Very good' else x)\
    .apply(lambda x: 5 if x == 'Excellent' else x)

<ipython-input-134-702b6ec16249>:1: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


df_case3 = df_case3.merge(health, on='mergeid')


df_case3 = df_case3.merge(df[['mergeid']].join(corr_df[['Factor_1', 'Factor_2']]), on='mergeid')


df_case3.dropna(inplace=True)


from sklearn.model_selection import train_test_split
X, y = df_case3[[i for i in df_case3.columns if i != 'mergeid' and i !='Factor_1' and  i != 'Factor_2']].values,\
df_case3['Factor_1'].values


from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)


from sklearn import linear_model
clf = linear_model.Lasso()


clf.fit(X_train, y_train)

Lasso()


pred = clf.predict(X)


clf.coef_

array([ 0., -0., -0.,  0.,  0.,  0.,  0.,  0.,  0., -0., -0.,  0., -0.,
       -0., -0.,  0.,  0., -0., -0., -0.,  0.,  0.,  0.,  0., -0., -0.,
       -0.,  0.,  0.,  0., -0.,  0.,  0., -0.,  0., -0.,  0.,  0., -0.,
        0., -0., -0.,  0., -0.,  0., -0.])


clf.score(X_test, y_test)

-3.8590865895304916e-05


X, y = df_case3[[i for i in df_case3.columns if i != 'mergeid' and i !='Factor_1' and  i != 'Factor_2']].values,\
            df_case3['Factor_2'].values


X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=15)


clf_2 = linear_model.Lasso()
clf_2.fit(X_train, y_train)

Lasso()


clf_2.coef_

array([-0.,  0.,  0., -0., -0., -0., -0., -0., -0.,  0.,  0., -0.,  0.,
       -0.,  0.,  0., -0.,  0.,  0.,  0., -0., -0., -0., -0.,  0.,  0.,
        0.,  0., -0., -0., -0., -0., -0.,  0., -0.,  0., -0., -0.,  0.,
       -0.,  0.,  0., -0.,  0., -0.,  0.])


clf_2.score(X_test, y_test)

-0.00024748253322748326


from sklearn.ensemble import RandomForestRegressor
rf = RandomForestRegressor(random_state=0)


X, y = df_case3[[i for i in df_case3.columns if i != 'mergeid' and i !='Factor_1' and  i != 'Factor_2']].values,\
df_case3['Factor_1'].values

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)


rf.fit(X_train, y_train)

RandomForestRegressor(random_state=0)


rf.score(X_test, y_test)

0.985843241063699


pred = rf.predict(X_test)


plt.plot(y_test, pred,'.')

[<matplotlib.lines.Line2D at 0x2622a3e30a0>]


pd.DataFrame({"Features":df_case3[[i for i in df_case3.columns if i != 'mergeid' and i !='Factor_1' and  i != 'Factor_2']].columns,
              "Feature Importance":rf.feature_importances_})


X, y = df_case3[[i for i in df_case3.columns if i != 'mergeid' and i !='Factor_1' and  i != 'Factor_2']].values,\
            df_case3['Factor_2'].values
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=15)


rf_2 = RandomForestRegressor(random_state=2)
rf_2.fit(X_train, y_train)

RandomForestRegressor(random_state=2)


rf_2.score(X_test, y_test)

0.9902248312949842


pred = rf_2.predict(X_test)
plt.plot(y_test, pred,'.')

[<matplotlib.lines.Line2D at 0x2622a43b2b0>]


pd.DataFrame({"Features":df_case3[[i for i in df_case3.columns if i != 'mergeid' and i !='Factor_1' and  i != 'Factor_2']].columns,
              "Feature Importance":rf_2.feature_importances_})

	Median Wage	Life Satisfaction
country
Austria	27600.000000	8.253930
Belgium	29557.368651	7.738783
Bulgaria	3681.357906	6.210499
Croatia	8038.908513	7.117545
Cyprus	18000.000000	7.897864
Czech Republic	10028.109012	7.599840
Denmark	43556.582857	8.536147
Estonia	9480.000000	6.888740
Finland	33600.000000	8.329060
France	27600.000000	7.450285
Germany	30000.000000	7.857905
Greece	12000.000000	7.049641
Hungary	7606.972999	6.874231
Israel	35448.770996	7.956146
Italy	20400.000000	7.578187
Latvia	6000.000000	6.624591
Lithuania	6000.000000	6.491956
Luxembourg	55181.852981	7.987478
Malta	12000.000000	8.167785
Poland	8456.659680	7.469256
Portugal	12000.000000	7.684457
Romania	3939.765389	7.367184
Slovakia	12000.000000	7.617378
Slovenia	14400.000000	7.424198
Spain	15600.000000	7.640291
Sweden	33627.049822	8.354010
Switzerland	53971.392346	8.432373

	MR1	MR2
country
Austria	0.325693	-0.457261
Belgium	0.201459	-0.122745
Bulgaria	-0.564852	0.433664
Croatia	-0.133694	0.219955
Cyprus	-0.094447	0.293831
Czech Republic	-0.170374	0.225735
Denmark	0.585978	-0.394890
Estonia	-0.300666	0.034354
Finland	0.222840	0.019432
France	0.167900	-0.067557
Germany	0.302761	-0.253279
Greece	-0.535937	0.610024
Hungary	-0.399798	0.025018
Israel	0.061645	0.123982
Italy	-0.154268	0.278524
Latvia	-0.536892	0.123840
Lithuania	-0.410271	0.608445
Luxembourg	0.370236	-0.416786
Malta	0.128402	-0.071670
Poland	0.003619	-0.056039
Portugal	-0.391569	0.208524
Romania	NaN	NaN
Slovakia	NaN	NaN
Slovenia	0.093621	-0.139435
Spain	-0.156529	-0.016771
Sweden	0.186603	-0.242022
Switzerland	0.513112	-0.385994

	var	varlab	values	uniques	nas
9	wq018_1	Work has heavy time pressure	double	7	72555
24	wq018_2	Work has heavy time pressure	double	6	72755
39	wq018_3	Work has heavy time pressure	double	7	74064
54	wq018_4	Work has heavy time pressure	double	7	75120
69	wq018_5	Work has heavy time pressure	double	7	75986
84	wq018_6	Work has heavy time pressure	double	6	76442
99	wq018_7	Work has heavy time pressure	double	7	76772
114	wq018_8	Work has heavy time pressure	double	6	76939
129	wq018_9	Work has heavy time pressure	double	6	77070
144	wq018_10	Work has heavy time pressure	double	6	77129
159	wq018_11	Work has heavy time pressure	double	5	77168
174	wq018_12	Work has heavy time pressure	double	5	77179
189	wq018_13	Work has heavy time pressure	double	5	77190
204	wq018_14	Work has heavy time pressure	double	5	77195
219	wq018_15	Work has heavy time pressure	double	4	77197
234	wq018_16	Work has heavy time pressure	double	2	77201
249	wq018_18	Work has heavy time pressure	double	2	77201
264	wq004_	Work had heavy time pressure	double	7	38765

	var	varlab	values	uniques	nas
10	wq019_1	Work is emotionally demanding	double	7	72555
25	wq019_2	Work is emotionally demanding	double	7	72755
40	wq019_3	Work is emotionally demanding	double	7	74064
55	wq019_4	Work is emotionally demanding	double	7	75120
70	wq019_5	Work is emotionally demanding	double	7	75986
85	wq019_6	Work is emotionally demanding	double	6	76442
100	wq019_7	Work is emotionally demanding	double	7	76772
115	wq019_8	Work is emotionally demanding	double	6	76939
130	wq019_9	Work is emotionally demanding	double	6	77070
145	wq019_10	Work is emotionally demanding	double	6	77129
160	wq019_11	Work is emotionally demanding	double	5	77168
175	wq019_12	Work is emotionally demanding	double	5	77179
190	wq019_13	Work is emotionally demanding	double	5	77190
205	wq019_14	Work is emotionally demanding	double	4	77195
220	wq019_15	Work is emotionally demanding	double	4	77197
235	wq019_16	Work is emotionally demanding	double	2	77201
250	wq019_18	Work is emotionally demanding	double	2	77201
265	wq005_	Work was emotionally demanding	double	7	38765

	var	varlab	values	uniques	nas
11	wq020_1	Work involves conflicts	double	7	72555
26	wq020_2	Work involves conflicts	double	7	72755
41	wq020_3	Work involves conflicts	double	7	74064
56	wq020_4	Work involves conflicts	double	7	75120
71	wq020_5	Work involves conflicts	double	7	75986
86	wq020_6	Work involves conflicts	double	6	76442
101	wq020_7	Work involves conflicts	double	6	76772
116	wq020_8	Work involves conflicts	double	5	76939
131	wq020_9	Work involves conflicts	double	6	77070
146	wq020_10	Work involves conflicts	double	6	77129
161	wq020_11	Work involves conflicts	double	5	77168
176	wq020_12	Work involves conflicts	double	5	77179
191	wq020_13	Work involves conflicts	double	5	77190
206	wq020_14	Work involves conflicts	double	5	77195
221	wq020_15	Work involves conflicts	double	3	77197
236	wq020_16	Work involves conflicts	double	2	77201
251	wq020_18	Work involves conflicts	double	2	77201
266	wq006_	Work involved conflicts	double	7	38765

Student Data:¶

My work:¶

Case 2¶

Setting up environment and loading the data¶

Formating data¶

Fixing NAs for Romania so we can keep it for the description part¶

Standardizing¶

Descriptive analysis¶

Factor Analysis¶

Case 3¶

Building poor_jq¶

Pressure¶

emo_demanding¶

conflicts¶

phys_demanding, uncomfortable¶

security (health risk reduced), freedom (will asume reverse values), skills, support, recognition¶

Employment¶

Socio-demographic factors¶

Health¶

Modelling¶

MR1¶

MR2¶

Trying with Random Forest¶

MR1¶

MR2¶

	ac014_	ac015_	ac016_	ac017_	ac018_	ac019_	ac020_	ac021_	ac022_	ac023_	ac024_	ac025_
ac014_	1.000000	0.452976	0.386468	-0.261215	0.140704	0.252529	-0.206355	-0.257534	-0.154932	-0.416880	-0.344952	-0.364573
ac015_	0.452976	1.000000	0.503004	-0.233042	0.198521	0.250057	-0.224599	-0.268606	-0.176017	-0.351900	-0.296373	-0.329618
ac016_	0.386468	0.503004	1.000000	-0.251771	0.206150	0.290144	-0.263379	-0.322571	-0.230004	-0.341738	-0.345645	-0.354189
ac017_	-0.261215	-0.233042	-0.251771	1.000000	-0.047013	-0.164002	0.281548	0.321422	0.232715	0.366042	0.348815	0.350667
ac018_	0.140704	0.198521	0.206150	-0.047013	1.000000	0.246160	-0.068939	-0.064999	-0.064887	-0.044516	-0.061381	-0.070154
ac019_	0.252529	0.250057	0.290144	-0.164002	0.246160	1.000000	-0.173317	-0.192027	-0.193867	-0.195230	-0.272121	-0.296930
ac020_	-0.206355	-0.224599	-0.263379	0.281548	-0.068939	-0.173317	1.000000	0.553536	0.361880	0.384247	0.413541	0.437354
ac021_	-0.257534	-0.268606	-0.322571	0.321422	-0.064999	-0.192027	0.553536	1.000000	0.450083	0.444719	0.474275	0.500598
ac022_	-0.154932	-0.176017	-0.230004	0.232715	-0.064887	-0.193867	0.361880	0.450083	1.000000	0.331900	0.383088	0.402166
ac023_	-0.416880	-0.351900	-0.341738	0.366042	-0.044516	-0.195230	0.384247	0.444719	0.331900	1.000000	0.562482	0.557246
ac024_	-0.344952	-0.296373	-0.345645	0.348815	-0.061381	-0.272121	0.413541	0.474275	0.383088	0.562482	1.000000	0.643846
ac025_	-0.364573	-0.329618	-0.354189	0.350667	-0.070154	-0.296930	0.437354	0.500598	0.402166	0.557246	0.643846	1.000000

	var	varlab	values	uniques	nas
262	wq002_	Work was physically demanding	double	7	38765
263	wq003_	Work was uncomfortable	double	7	38765
264	wq004_	Work had heavy time pressure	double	7	38765
265	wq005_	Work was emotionally demanding	double	7	38765
266	wq006_	Work involved conflicts	double	7	38765
267	wq007_	Work had little freedom to decide	double	7	38765
268	wq008_	Work allowed development of skills	double	7	38765
269	wq009_	Work gave recognition	double	7	38766

	var	varlab	values	uniques	nas
15	wq024_1	Work has adequate salary	double	7	72555
30	wq024_2	Work has adequate salary	double	7	72755
45	wq024_3	Work has adequate salary	double	7	74064
60	wq024_4	Work has adequate salary	double	7	75120
75	wq024_5	Work has adequate salary	double	7	75986
90	wq024_6	Work has adequate salary	double	6	76442
105	wq024_7	Work has adequate salary	double	7	76772
120	wq024_8	Work has adequate salary	double	6	76939
135	wq024_9	Work has adequate salary	double	6	77070
150	wq024_10	Work has adequate salary	double	6	77129
165	wq024_11	Work has adequate salary	double	5	77168
180	wq024_12	Work has adequate salary	double	5	77179
195	wq024_13	Work has adequate salary	double	3	77190
210	wq024_14	Work has adequate salary	double	3	77195
225	wq024_15	Work has adequate salary	double	5	77197
240	wq024_16	Work has adequate salary	double	2	77201
255	wq024_18	Work has adequate salary	double	2	77201
270	wq010_	Work had adequate salary	double	7	38765

	mergeid	wq010_
0	AT-001215-01	2.0
2	AT-001492-02	4.0
3	AT-001881-01	3.0
4	AT-001881-02	3.0
5	AT-002136-01	3.0
...	...	...
77196	SK-993822-02	3.0
77197	SK-995042-01	4.0
77198	SK-995042-02	4.0
77200	SK-999958-01	2.0
77201	SK-999958-02	3.0

	var	varlab	values	uniques	nas
15	ep018_	Kind of industry working in	double	1	32949
34	ep054_	Kind of industry working in last job	double	1	32949

	var	varlab	values	uniques	nas
43	re014_1	Job industry	double	17	17962
59	re014_2	Job industry	double	17	37330
75	re014_3	Job industry	double	17	53139
91	re014_4	Job industry	double	17	63405
107	re014_5	Job industry	double	17	69561
123	re014_6	Job industry	double	17	72925
139	re014_7	Job industry	double	16	74843
155	re014_8	Job industry	double	17	75932
171	re014_9	Job industry	double	16	76553
187	re014_10	Job industry	double	16	76866
203	re014_11	Job industry	double	15	77023
219	re014_12	Job industry	double	14	77096
232	re014_13	Job industry	double	15	77143
246	re014_14	Job industry	double	13	77168
262	re014_15	Job industry	double	9	77182
274	re014_16	Job industry	double	7	77191
286	re014_17	Job industry	double	6	77195
295	re014_18	Job industry	double	4	77197
307	re014_19	Job industry	double	5	77198
316	re014_20	Job industry	double	4	77199

	var	varlab	values	uniques	nas
44	re015_1	Was employee civil servant or self	double	6	17962
48	re023_1	First monthly work income in self-employment	double	136	75284
51	re027_1	Current wage if still employed	double	383	72936
53	re029_1	Current work income if still self-employed	double	61	76824
60	re015_2	Was employee civil servant or self	double	6	37330
...	...	...	...	...	...
287	re015_17	Was employee civil servant or self	double	3	77195
296	re015_18	Was employee civil servant or self	double	3	77197
301	re027_18	Current wage if still employed	double	77202	77201
308	re015_19	Was employee civil servant or self	double	3	77198
317	re015_20	Was employee civil servant or self	double	3	77199

	var	varlab	values	uniques	nas
6	re002_	Year finished fulltime education	double	100	14113
45	re016_1	Job was part or full time	double	8	17962
61	re016_2	Job was part or full time	double	8	37330
77	re016_3	Job was part or full time	double	8	53139
93	re016_4	Job was part or full time	double	8	63405
109	re016_5	Job was part or full time	double	8	69561
125	re016_6	Job was part or full time	double	7	72925
141	re016_7	Job was part or full time	double	7	74843
157	re016_8	Job was part or full time	double	8	75932
173	re016_9	Job was part or full time	double	6	76553
189	re016_10	Job was part or full time	double	7	76866
205	re016_11	Job was part or full time	double	6	77023
221	re016_12	Job was part or full time	double	5	77096
234	re016_13	Job was part or full time	double	5	77143
248	re016_14	Job was part or full time	double	4	77168
264	re016_15	Job was part or full time	double	3	77182
276	re016_16	Job was part or full time	double	3	77191
288	re016_17	Job was part or full time	double	2	77195
297	re016_18	Job was part or full time	double	2	77197
309	re016_19	Job was part or full time	double	3	77198
318	re016_20	Job was part or full time	double	2	77199

	age	gender	single	mstat	yedu	cjs	thinc2
mergeid
AT-001215-01	78.0	Female	Yes	Never married	15.0	Retired	14400.0
AT-001492-01	65.0	Female	No	Married, living with spouse	11.0	Homemaker	24000.0
AT-001492-02	66.0	Male	No	Married, living with spouse	13.0	Retired	24000.0
AT-001881-01	87.0	Female	Yes	Widowed	12.0	Retired	17400.0
AT-001881-02	93.0	Male	Yes	Married, living with spouse	3.0	Retired	15600.0
...	...	...	...	...	...	...	...
SK-995042-01	59.0	Female	No	Married, living with spouse	10.0	Retired	15600.0
SK-995042-02	62.0	Male	No	Married, living with spouse	10.0	Retired	15600.0
SK-996004-01	56.0	Female	Yes	Divorced	13.0	Employed or self-employed	10596.0
SK-999958-01	71.0	Male	No	Married, living with spouse	11.0	Retired	10800.0
SK-999958-02	72.0	Female	No	Married, living with spouse	11.0	Retired	10800.0

	thinc2
count	142831.000000
mean	26393.808218
std	26502.360081
min	0.000000
33%	12000.000000
50%	18000.000000
66%	27600.000000
max	403301.693123

	Features	Feature Importance
0	wq010_	0.045297
1	poor_jq	0.156539
2	bad_stress	0.142503
3	ac035d1	0.012429
4	ac035d4	0.008372
5	ac035d7	0.005430
6	ac035d8	0.015287
7	ac035d9	0.019444
8	ac035d10	0.015679
9	ac035dno	0.020574
10	type_of_emp_Civil servant (public sector)	0.014248
11	type_of_emp_Employee (private sector)	0.011756
12	type_of_emp_Self-employed (including working f...	0.004019
13	full_part_Always full-time	0.005258
14	full_part_Always part-time	0.005569
15	full_part_Changed multiple times	0.002811
16	full_part_Changed once from full-time to part-...	0.003818
17	full_part_Changed once from part-time to full-...	0.001162
18	industry_Agriculture, hunting, forestry, fishing	0.013933
19	industry_Construction	0.014088
20	industry_Education	0.006266
21	industry_Electricity, gas and water supply	0.005208
22	industry_Financial intermediation	0.003351
23	industry_Health and social work	0.008285
24	industry_Hotels and restaurants	0.010073
25	industry_Manufacturing	0.020435
26	industry_Mining and quarrying	0.004760
27	industry_Other community	0.013255
28	industry_Public administration and defence	0.008006
29	industry_Real estate, renting and business act...	0.001080
30	industry_Transport, storage and communication	0.011074
31	industry_Wholesale and retail trade	0.015007
32	Female	0.011792
33	Male	0.011751
34	edu_cat_edu_high	0.009663
35	edu_cat_edu_low	0.014610
36	edu_cat_edu_med	0.016202
37	Income_cat_income_high	0.023604
38	Income_cat_income_low	0.020210
39	Income_cat_income_med	0.008405

	Features	Feature Importance
0	wq010_	0.049591
1	poor_jq	0.173799
2	bad_stress	0.142549
3	ac035d1	0.013811
4	ac035d4	0.008390
5	ac035d7	0.007451
6	ac035d8	0.015849
7	ac035d9	0.019101
8	ac035d10	0.017099
9	ac035dno	0.011959
10	type_of_emp_Civil servant (public sector)	0.011984
11	type_of_emp_Employee (private sector)	0.012117
12	type_of_emp_Self-employed (including working f...	0.005328
13	full_part_Always full-time	0.005906
14	full_part_Always part-time	0.004883
15	full_part_Changed multiple times	0.002565
16	full_part_Changed once from full-time to part-...	0.003099
17	full_part_Changed once from part-time to full-...	0.001907
18	industry_Agriculture, hunting, forestry, fishing	0.013861
19	industry_Construction	0.012283
20	industry_Education	0.008587
21	industry_Electricity, gas and water supply	0.004085
22	industry_Financial intermediation	0.004859
23	industry_Health and social work	0.009912
24	industry_Hotels and restaurants	0.009256
25	industry_Manufacturing	0.020988
26	industry_Mining and quarrying	0.005739
27	industry_Other community	0.013880
28	industry_Public administration and defence	0.007714
29	industry_Real estate, renting and business act...	0.001312
30	industry_Transport, storage and communication	0.008238
31	industry_Wholesale and retail trade	0.016771
32	Female	0.012960
33	Male	0.012834
34	edu_cat_edu_high	0.010787
35	edu_cat_edu_low	0.015141
36	edu_cat_edu_med	0.015246
37	Income_cat_income_high	0.012171
38	Income_cat_income_low	0.014739
39	Income_cat_income_med	0.010061

	var	varlab	values	uniques	nas
13	chronic2w7	2+ chronic diseases (w7 version)	double	3	536
14	chronicw7c	Number of chronic diseases (w7 version)	double	550	536